Can Yi 34B run on the H100 SXM?
Yes. At 8K context the best fit is F16, using about 66.5 GB of the 76.0 GB usable on the H100 SXM. Estimated speed ceiling: ~49 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | Max tok/s (est., 8K) |
|---|---|---|
| 2-bit (Q2_K) • | 15.0 GB Fits | 230 |
| 3-bit (Q3_K_M) • | 17.0 GB Fits | 201 |
| 4-bit (Q4_K_M) • | 20.7 GB Fits | 162 |
| 5-bit (Q5_K_M) • | 24.1 GB Fits | 138 |
| 6-bit (Q6_K) • | 27.7 GB Fits | 119 |
| 8-bit (Q8_0) • | 35.5 GB Fits | 92 |
| 16-bit (F16/BF16) | 65.5 GB Fits | 49 |
• real GGUF file size from TheBloke/Yi-34B-GGUF. Speed is a bandwidth ceiling (3350 GB/s), not a benchmark. Methodology.
More
Yi 34B on every GPU and Mac · Every model that runs on the H100 SXM · Other models that fit 64 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (01-ai/Yi-34B), NVIDIA specs. Data as of 2026-11-17.